Close This Window

Forest Ecosystem Dynamics Multisensor Airborne Campaign

Project Description
The Biospheric Sciences Branch (formerly Earth Resources Branch)
within the Laboratory for Terrestrial Physics at NASA's Goddard Space
Flight Center and associated University investigators are involved in
a research program entitled Forest Ecosystem Dynamics (FED) which is
fundamentally concerned with vegetation change of forest ecosystems at
local to regional spatial scales (100 to 10,000 meters) and temporal
scales ranging from monthly to decadal periods (10 to 100 years). The
nature and extent of the impacts of these changes, as well as the
feedbacks to global climate, may be addressed through modeling the
interactions of the vegetation, soil, and energy components of the
boreal ecosystem.

The FED ecosystem modeling research efforts concentrate on the North
American boreal and northern hardwood transition forests with
emphasis on optical and radar remote sensing technology. This
research employs an integrated approach of field and aircraft
studies, theoretical modeling, and satellite image data processing to
infer where landscape pattern and process ecosystem model predictions
succeed or fail at regional spatial scales and interannual temporal
scales. We are also using remote sensing observations as a check on
potentially observable forest ecosystem model predicted attributes
(e.g., species composition, tree height distributions, land use
patterns). Conversely, we are investigating the potential of remote
sensing observations for extracting biophysical properties of forest
canopies, soils, and hydrologic parameters used in our forest
ecosystem models. On-going work, in addition to activities discussed
here, include modeling, measurement, and data compilation or a number of
boreal zone sites.

The FED model framework (see Levine et al., 1983) integrates existing
models of forest growth and succession (i.e., FORET model of Shugart
and West, 1984 and ZELIG Model of Smith and Urban, 1988), soil
processes (Residue model of Bidlake et al., 1992; Bristow, et al.,
1986; TERRA model of Levine, 1984; Levine and Ciolkosz, 1988), and
energy dynamics (e.g., Smith et. al., 1981; Kimes and Kirchner,
1982). Each of the models interact with the others to provide feedback
controls on growth, soil related processes, and energy internal and
external to the forest environment. The forest succession and soil
process models require input at the species and soil characteristics
level, respectively. This makes this formulation useful for examining
the effects of changes in climate or anthropogenic factors on the
community composition and structure of the boreal forest.

The results anticipated from this experiment will enable development
and validation of the integrated model to usefully characterize the
ecosystem dynamics of the boreal forest under a variety of
conditions. A number of questions pertinent to the combined
experiment may then be considered. For example, how do climatic
gradients determine the spatial distribution of species within the
boreal forest? What are the possible effects of global climate change
on the boreal forest? Is the boreal forest a net source or sink of
carbon and methane and will the present state change if climate
changes? Also relevant to the issue of global change are the
magnitudes of the feedbacks between climate and vegetation. The
model, as formulated, can provide insights into the effects of
climate change on ecosystem dynamics, but does not consider the
effects of ecosystem changes on climate directly. However, the model
can provide, as outputs, factors that impact climate such as albedo,
evapotranspiration, and trace gas fluxes (i.e., carbon dioxide,
methane, and nitrogen). These questions are also relevant to the
BOREAS experiment.

The overall objective of our work was to capitalize on and develop the
unique advantages of remote sensing data combined with models of
forest ecosystem dynamics for characterizing northern/boreal forest
ecosystems, especially with regard to the interpretation of landscape
patterns and processes at local and regional scales. Specific
objectives for the FED experiment at IP's Northern Experimental Forest
included:

1. Enhance the development of an integrated quantitative model which
simulates forest, soil, and energy dynamics processes in northern
forest environments. This will be achieved through modification and
continuing development of the three types of sub-models discussed
above.

2. Develop improved remote sensing technology to infer biophysical
parameter inputs for forest succession and soil models. The
relationships among remote sensing and forest canopy
characteristics required by forest succession and soil process
models will be developed and tested.

3. Develop a better understanding of the transfer and utilization of
energy in forest canopies. This goal is being accomplished via
collection of detailed spectral reflectance data in the field and
laboratory, and by exercising existing radiative transfer models.

4. Use field and remote sensing observations to help infer where
landscape pattern and process ecosystem models succeed or fail at
local to regional spatial scales and interannual temporal
scales. This objective is being accomplished through comparison of
model predictions with field experimental data, and changes in
successional stage, bioproductivity, and other biophysical
parameters based on remotely sensed measurements.

5. Use field and remote sensing observations to help infer where
landscape pattern and process ecosystem models succeed or fail at
local to regional spatial scales and interannual temporal
scales. This objective is being accomplished through comparison of
model predictions with field experimental data, and changes in
successional stage, bioproductivity, and other biophysical
parameters based on remotely sensed measurements.

6. Use remote sensing observations as a check on such potentially
observable forest ecosystem dynamic model predicted
attributes. Specific ecosystem model algorithms and output
parameters are being evaluated directly by examining relationships
developed between sensor measured response and ecosystem
attributes.

7. Use remote sensing observations and models to extract biophysical
properties of forest canopies, soils, and hydrologic parameters
used in our forest ecosystem models. Physically based radar and
optical models are being applied to data collected over subsets of
the Northern Experimental Forest to examine radar and optical
scattering characteristics of different scene components. Model
inversion strategies are also being applied for selected ecosystem
model inputs.

For more information, link to
http://forest.gsfc.nasa.gov/html/fedmac/fedmac.html
Close This Window